Article Text

Download PDFPDF

Original research
A comparative analysis of 3MAX aspiration versus 3 mm Trevo Retriever for distal occlusion thrombectomy in acute stroke
  1. Diogo C Haussen1,
  2. Brendan Eby,
  3. Alhamza R Al-Bayati2,
  4. Jonathan A Grossberg3,
  5. Gabriel Martins Rodrigues,
  6. Michael R Frankel4,5,
  7. Raul G Nogueira
  1. 1 Neurology, Neurosurgery and Radiology, Emory University School of Medicine/Marcus Stroke & Neuroscience Center - Grady Memorial Hospital, Atlanta, Georgia, USA
  2. 2 Marcus Stroke & Neuroscience Center - Grady Memorial Hospital, Emory University School of Medicine, Atlanta, Georgia, USA
  3. 3 Neurosurgery and Radiology, Emory University School of Medicine, Atlanta, Georgia, USA
  4. 4 Department of Neurology, Emory University School of Medicine, Atlanta, Georgia, USA
  5. 5 Marcus Stroke and Neuroscience Center, Grady Memorial Hospital, Atlanta, Georgia, USA
  1. Correspondence to Dr. Raul G Nogueira; raul.g.nogueira{at}


Background Although aspiration and stent retriever thrombectomy perform similarly in proximal occlusions, no comparative series are available in distal occlusions. We aimed to compare the 3 mm Trevo Retriever against the 3MAX thromboaspiration catheter in distal arterial occlusions.

Methods A single-center retrospective review of a prospectively maintained databank for patients treated with the 3 mm Trevo stent retriever or 3MAX thromboaspiration as the upfront approach for distal occlusions (middle cerebral artery mid/distal M2/M3, anterior cerebral artery A1/A2/A3 or posterior cerebral artery P1/P2) from January 2014 to July 2018 was performed. The primary outcome was the rate of distal occlusion first-pass reperfusion (modified Thrombolysis in Cerebral Infarction (mTICI) 2b–3).

Results Of 1100 patients treated within the study period, 137 patients/144 different arteries were treated with the 3 mm Trevo (n=92) or 3MAX device (n=52). The groups had comparable demographics and baseline characteristics. There was a higher rate of first-pass mTICI 2b–3 reperfusion (62% vs 44%; p=0.03), a trend towards a higher rate of final mTICI 2b–3 reperfusion (84% vs 69%; p=0.05), and lower use of adjuvant therapy (15% vs 31%; p=0.03) with the 3 mm Trevo compared with the 3MAX. The median number of passes (p=0.46), frequency of arterial spasm (p=1.00), rates of parenchymal hematomas (p=0.22)/subarachnoid hemorrhage (p=0.37) in the territory of the approached vessel were similar across the two groups. The 90-day rate of good outcomes (45% vs 46% in the 3 mm Trevo and 3MAX groups, respectively; p=0.84) was comparable. Multivariable regression identified baseline NIH Stroke Scale (NIHSS) score (OR 0.9; 95% CI 0.8 to 0.97; p<0.01) and use of 3 mm Trevo (OR 2.2; 95% CI 1.1 to 4.6; p=0.02) independently associated with first-pass mTICI 2b–3 reperfusion.

Conclusions In the setting of distal arterial occlusions, the 3 mm Trevo may lead to higher rates of first-pass reperfusion than direct 3MAX thromboaspiration. Lower NIHSS was found to be associated with improved reperfusion rates as observed in more proximal lesions. Further studies are warranted.

  • stroke
  • thrombectomy
  • stent

Statistics from


Distal occlusion strokes, defined as middle cerebral artery (MCA)-M2, MCA-M3, anterior cerebral artery (ACA), or posterior cerebral artery (PCA) occlusions, represent approximately 40% of patients with acute ischemic stroke (AIS) with visible occlusions on CT angiography.1

Intravenous thrombolysis is clinically more effective in patients with visible occlusions, with higher rates of reperfusion in more distal vessels, presumably as a function of lower clot burden. However, distal occlusions may still lead to significant long-term disability even in the setting of IV recombinant tissue plasminogen activator treatment. Indeed, 40% of patients with an M2 occlusion were dead at 90 days in the Interventional Management of Stroke (IMS) III trial.2

The available controlled data for the efficacy of thrombectomy in distal occlusions is limited to M2 occlusions. The HERMES meta-analysis of randomized clinical trials in thrombectomy for stroke showed that MCA M2 thrombectomy may be helpful,3 and a multicenter retrospective study found improved outcomes in individuals with M2 occlusions treated with thrombectomy.4

Although aspiration and stent retriever mechanical thrombectomy (MT) perform similarly in proximal occlusions, data on more distal occlusions are limited.5 6 The 3×20 mm Trevo Retriever (3 mm Trevo) and the Penumbra 3MAX reperfusion catheter have been reported in distal occlusions with good reperfusion rates and low safety concerns.7–13 However, the optimal approach for distal intracranial occlusions in thrombectomy for AIS is unclear since no comparative data are available. In this study we aim to compare our experience with the 3 mm Trevo Retriever against 3MAX thromboaspiration in distal occlusions.


This is a retrospective review of a prospectively collected thrombectomy databank for consecutive cases of use of 3 mm Trevo (Stryker, Freemont, California, USA) (delivered via a 2F outer diameter tip/0.017 inch inner diameter microcatheter) or 3MAX (Penumbra, Alameda, California, USA) (3.8F outer diameter tip/0.035 inch inner diameter) for distal arterial occlusions (DAO) involving the MCA (mid or distal M2 segment, M3 segment), ACA (A1, A2, A3), or PCA (P1, P2) from January 2014 to July 2018. This study was approved by the local Institutional Review Board with a waiver of consent.

The 3MAX thromboaspiration catheter was utilized using a modified ADAPT technique in which the microwire is not used to cross the thrombus unless the anatomy was unfavorable for safe 3MAX advancement into the face of the clot without wire purchase past the occlusion.14 The 3 mm Trevo was used as previously described.7 The device choice was made according to the operator’s preference.

Distal arterial occlusion (DAO) definitions

The occlusion site was defined according to the following criteria. ACA A1: from origin to the anterior communicating artery; A2: from the anterior communicating artery to the callosomarginal takeoff; A3: pericallosal or callosomarginal; PCA P1: from origin to the posterior communicating artery; P2+: distal to the posterior communicating artery; MCA mid M2: the distal half (midway or distal) of the vertical segment of the first MCA branch from the MCA limen insulae and circular sulcus on anterior–posterior angiography4; distal M2: immediately proximal to or at the M2–3 junction (including occlusions at the circular sulcus); M3: the opercular segment between the circular sulcus of the insula and the cortical surface.15

Angiograms were evaluated for the presence of a distal occlusion (modified Thrombolysis in Cerebral Infarction (mTICI) 0–1) and classified according to their clinical scenario: isolated distal occlusion; distal occlusion in the setting of a concomitant larger vessel/proximal occlusion; emboli to the same or new territory after thrombectomy of a proximal occlusion.

The diameter of the vessel immediately was calculated just proximal to the occlusion site via iSite PACS on cine images from fluoroscopic runs obtained on Phillips Allura biplane machine (Andover, Massachusetts, USA) in all cases. In stent retriever cases, additional measurements of the vessel diameter distal to the retriever tines were obtained.


The primary outcome was the rate of distal occlusion first-pass reperfusion (defined by mTICI 2b–3). Secondary efficacy endpoints included the number of passes attempted with the primary device, use of adjunct strategy employed (defined as the use of any other thrombectomy or thrombolysis technique due to failure of the upfront strategy), final degree of reperfusion, and the rates of good outcome (90-day modified Rankin Scale (mRS) 0–2 via in-person visit or structured telephone interview).16

Safety outcomes were recorded including the presence of hemorrhagic transformation, any parenchymal hematoma (ECASS III parenchymal hematoma type 1 or 2), and sulcal or subarachnoid hemorrhage (diffuse or in the territory of the artery treated for distal occlusion); hemorrhage identified in cases where the small device was used in the same territory of a treated proximal occlusion (emboli to same territory) was included even if a definitive causative relationship with the device could not be determined. The frequency of arterial spasm (defined as any degree of spasm in treated vessels) was recorded. Reperfusion rates were defined by the treating neurointerventionists and retrospectively re-checked by a fellowship-trained neurointerventionist who also rated hemorrhagic complications.

Statistical analysis

Continuous variables are reported as mean±SD or median (IQR). Categorical variables are reported as proportions. Between-group comparisons were made with the Student t-test/Mann–Whitney U test or χ2/Fisher exact test. Significance was set at p<0.05. Multivariate logistic regression analyses were performed for variables at the p<0.1 level of significance on univariate analysis. Analyses were performed using SPSS Statistics 24 (IBM, Armonk, New York, USA).


Baseline characteristics

Of 1100 patients treated with MT within the study period, 137 patients/144 different arteries underwent treatment with small devices (3 mm Trevo (n=92) vs 3MAX (n=52)) for DAO. The two groups were comparable in terms of demographics, use of intravenous thrombolysis, clinical severity on presentation, and baseline ischemic core and area of tissue at risk on CT perfusion (table 1). Moreover, there was no difference in the DAO presentation: isolated distal occlusion, emboli to new or same distal territory, or multiple occlusions (p=0.10). Technical feasibility (clot engagement) was 100% for both devices.

Table 1

Baseline characteristics

The occlusion site was comparable among groups (table 2). The approached distal vessel pre-clot diameter was larger by one-fifth of a millimeter in the stent retriever group compared with the aspiration arm (1.6 vs 1.4 mm; p=0.02).

Table 2

Distal occlusion site characteristics

Procedural outcomes

Table 3 shows the procedural, radiological, and clinical outcomes. There was a higher rate of first-pass mTICI 2b–3 reperfusion (62% vs 44%; p=0.03) and a trend towards a higher rate of final mTICI 2b–3 reperfusion (84% vs 69%; p=0.05) with the 3 mm Trevo compared with the 3MAX device. In addition, the use of adjuvant therapy was lower in the 3 mm Trevo group (15% vs 31%; p=0.03)(see online Supplemental Material).

Table 3

Procedural, radiological, and clinical outcomes

There were no differences in terms of first-pass (43% vs 34%; p=0.37) or final (58% vs 46%; p=0.22) mTICI 3 reperfusion across the two groups. The median number of passes (p=0.46) and the frequency of arterial spasm (p=1.00) were also comparable between groups (table 3).

Regarding the performance of the devices in specific arteries, the benefit of the stent retriever over thromboaspiration appeared more conspicuous in the ACA than in the MCA and PCA (see online supplementary e-Table). There was no difference between devices when evaluating arterial segments or divisions (see online supplementary e-Table).

Clinical/safety outcomes

The rates of parenchymal hematomas (p=0.22) and subarachnoid hemorrhage (p=0.37) in the territory of the approached vessel were similar across the two groups (table 3). When cases using adjuvant therapies were excluded (reperfusion attempts made with primary devices only), the frequency of subarachnoid hemorrhage was comparable with the 3 mm Trevo and the 3MAX device (2.6% vs 2.9%, respectively; p=0.92). The 90-day rate of good outcomes (45% vs 46% in the 3 mm Trevo and 3MAX groups, respectively; p=0.84) was comparable.

Predictors of first-pass reperfusion

Multivariable regression identified only baseline NIHSS (OR 0.9; 95% CI 0.8 to 0.97; p<0.01) and use of the 3 mm Trevo (OR 2.2; 95% CI 1.1 to 4.6; p=0.02) independently associated with first-pass mTICI 2b–3 reperfusion of distal occlusions.


In this cohort of consecutively treated patients undergoing emergent thrombectomy for AIS, better technical results were observed with the 3 mm Trevo stent retriever than with the Penumbra 3MAX device, as demonstrated by higher rates of first-pass mTICI 2b–3 reperfusion and lower utilization of adjuvant therapy. The association between first-pass reperfusion and the 3 mm Trevo remained present after adjustments for potential confounders. The overall safety profile was acceptable for both treatment approaches.

While first-pass reperfusion has been associated with less disability,17 we were unable to show a difference in clinical outcome. This presumably relates to the heterogeneity of tissue at-risk eloquence across subjects, variability of volume of tissue at risk, and type of occlusion at presentation (isolated distal occlusion vs distal occlusion in the setting of a concomitant larger vessel/proximal occlusion vs emboli to the same or new territory).7–13 18 In proximal occlusions, collateral strength has been demonstrated to be inversely associated with NIHSS and directly associated with improved reperfusion rates.19 We showed that lower NIHSS was independently associated with higher chances of first-pass reperfusion in patients with distal occlusions, which had not been previously shown. This may indicate that improved collaterals (lower NIHSS in the setting of small ischemic cores) can be associated with improved chances of reperfusion in distal segments.

The selection of the intra-arterial therapy approach is typically based on the expected branching pattern, vessel course (angulation), and caliber. However, there were no previous comparative studies regarding the safety and efficacy between stent retrievers and thromboaspiration for specific arterial sites, segments, or divisions in distal occlusions. The first report of the use of small retrievers (3×20 mm Trevo) was published in 2016. It indicated that stent retriever mechanical thrombectomy of distal occlusions (MCA M3, ACA A3, PCA P2 and P3) could safely lead to high rates of successful reperfusion (TICI 2b–3 in 75% of cases).7 Another study indicated that distal ACA occlusions (A2 or A3) could be treated with retrievers/retriever-like devices (in this study, only 90% of the devices were ≥4 mm diameter). In 16% of the cases the retriever could not be deployed and first-pass TICI 2b–3 reperfusion was achieved in 56% of all approached vessels. No parenchymal or subarachnoid hemorrhages were detected (13% cases of subarachnoid hemorrhage were seen but considered related to other vessel territory), but a distal ACA dissection occurred. The first-pass reperfusion rates for distal ACA was slightly lower (68% for the occlusions that could be engaged compared with 78% in the present study with 3 mm Trevo), which may relate to the smaller cell for clot capture of larger retrievers in sub-nominal vessels and the difficulty with distal access related to the larger microcatheter used for delivery of larger retrievers.8 Another series of 41 patients in which the Catch Mini retriever (BALT, Irvine, California, USA) was used in distal vessels (including M2, M3, ACA, and PCA) reported a final mTICI 2b–3 reperfusion rate of 78%. The highest recanalization rate was for M2 (82%), and one patient (2%) developed a parenchymal hematoma.10 This is similar to our 85% rate of final mTICI 2b–3 with the 3 mm Trevo. Our rate of parenchymal hematoma (6%) and subarachnoid hemorrhage (7%) was numerically higher. This could potentially relate to the fact that we did not include proximal M2 occlusions and frequently treated superior division MCA occlusions.

The theoretical benefit of thromboaspiration includes the potential to directly pushing the catheter into the proximal face of clot. This avoids crossing the thrombus with the microwire, therefore minimizing the risk of blind microwire advancement across the occlusion. However, it is not uncommon for the angle of engagement between the clot and the catheter to be excessive (such as a superior MCA M2–3 junction occlusion) leading to an unfavorable vector of engagement and expected suboptimal integration. This can be improved by crossing the occlusion with the microwire and maintaining it until clot contact, but may lead to vessel straightening and torsion, which is an undesirable feature of retrievers. A report of six patients treated for distal occlusions with the 3MAX showed 100% TICI 2b–3 reperfusion and no parenchymal hemorrhages.9 Another favorable study encompassed predominantly unselected MCA-M2 occlusions in which nearly all patients were pretreated with intravenous thrombolysis and treated with thrombectomy within 2 hours of onset.12 A report on the use of 3MAX in 38 distal occlusions (60% M2s) treated with 3MAX demonstrated a 59% rate of TICI 2b–3. Rescue 3×20 mm Trevo retriever was used in 42% of cases, comparing well with the 31% in the present series. Final successful reperfusion (TICI 2b–3) reached 76% and one vessel perforation occurred.11 Lastly, another study on 18 patients treated with 3MAX for distal occlusions led to final mTICI 2b–3 of 75%.13

This study has the limitations inherent to all retrospective analyses. Despite being the largest series on distal occlusions, the sample size is still small. Although the vessel diameter of the 3 mm Trevo group was slightly larger than the 3MAX group, the increased first-pass reperfusion in the retriever arm was present after adjustments. The reperfusion grades were self-adjudicated, but this applied to both devices. The 3MAX had been used in our institution prior to the study period while the 3 mm Trevo learning curve was included in this analysis. The subgroup analyses had limited power and therefore are prone to artifacts. Thrombectomy for distal occlusions is a controversial approach.


In the setting of distal arterial occlusions, the use of 3 mm stent retrievers may lead to higher rates of first-pass reperfusion compared with direct 3MAX thromboaspiration.

Lower NIHSS in the setting of distal occlusions was found to be associated with improved reperfusion rates, as observed in more proximal lesions. Further studies are warranted.



  • Contributors DCH: Study conception, design of the work, acquisition of data, statistical analysis, interpretation of data, drafting of the manuscript. BE, ARA, JAG, GMR, MRF: Data acquisition, critical revision of manuscript. RN: Design of the work, acquisition of data, interpretation of data, critical revision of manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests DCH: Consultant for Stryker and Vesalio. RN: Principal Investigator, Stryker Neurovascular (DAWN trial (no compensation), Trevo‑2 trial), Cerenovus/Neuravi (ENDOLOW trial, no compensation); consultant to Stryker Neurovascular; steering committee member, Stryker Neurovascular (no compensation), Medtronic (SWIFT trial, SWIFT Prime trial (no compensation)), Cerenovus/Neuravi (ARISE‑2 trial, no compensation); angiographic core lab, Medtronic (STAR trial); executive committee mem­ber, Penumbra (no compensation); physician advi­sory board, Cerenovus/Neuravi, Phenox, Anaconda, Genentech, Biogen, Prolong Pharmaceuticals, Allm (no compensation), Viz-AI; stock options, Viz-AI.

  • Ethics approval Emory University Institutions Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement The unpublished data from this dataset is held by Grady Memorial Hospital/Emory University and DCH/RN. Requests for data sharing would be required to be discussed with them directly.

  • Patient consent for publication Not required.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.